class: inverse, left, bottom, hide_logo background-image: url("images/rso.png") background-position: 10% 10%; background-size: 50px # Introduction to Weibull Analysis ## 30 Jan 2022
--- ## Test icon - This should be a finger pointing up (bouncing) <span class="faa-bounce animated" style="display: -moz-inline-stack; display: inline-block; transform: rotate(0deg);"><img src="☝" style="height:1em; width:auto;"></span> - <span class="faa-bounce animated" style="display: -moz-inline-stack; display:inline-block; transform rotate(0deg);"><img src="☝" style="height:1em; width: auto;"></span> -
<img src="☝" style="height:1em; width:auto; "/>
-
<img src="<U+261D>" style="height:1em; width:auto; "/>
- <span class="faa-bounce animated" style="display: -moz-inline-stack; display: inline-block; transform: rotate(0deg);"><img src="\u261d" style="height:1em; width:auto;"></span> -
-
-
<span id="MathJax-Element-53-Frame" class="mjx-chtml MathJax_CHTML">T</span> <math xmlns="http://www.w3.org/1998/Math/MathML">F</math> --- ## Distribution Functions - PDF & CDF .panelset[ .panel[.panel-name[Probability Density Function - PDF] .pull-left[
] .pull-right[ The probability density function (aka density function, or just density) defines the probability that the R.V. `\(T\)` is equal to `\(t\)`. $$ f(t) = \Pr(T = t) $$ The PDF for the Weibull distribution is expressed as $$ f(t) = \frac{\beta}{\eta}\left(\frac{t}{\eta}\right)^{\beta-1}\exp\left[-\left(\frac{t}{\eta}\right)^{\beta}\right] $$ ] <!--end .pull-right --> ] .panel[.panel-name[Cumulative Distribution Function - CDF] .pull-left[
] .pull-right[ The CDF is formally defined as the cumulative probability that the R.V. <span class="mjx-char MJXc-TeX-math-I" style="display:inline;">T</span> is equal to or less than <span class="mjx-char MJXc-TeX-math-I" style="display:inline;">t</span> $$ F(t) = \Pr(T \le t) $$ Less formally, `\(F(t)\)` is equal to the area under the curve of the density function in the interval `\([0,t]\)`, i.e. $$ F(t) = \int_{0}^{t}f(u)du $$ The Weibull CDF is expressed as $$ F(t) = 1-\exp\left[-\left(\frac{t}{\eta}\right)^{\beta}\right] $$ ] <!--end .pull-right --> ] <!--end .panel --> .panel[.panel-name[CDF/PDF Relationship] <img src="data:image/png;base64,#images/cdf_pdf.gif" style="display: block; margin: auto;" /> ] ] <!--end .panelset --> --- ## stuff - <span class="mjx-char MJXc-TeX-math-I" style="padding-top: 0.472em; padding-bottom: 0.259em; padding-right: 0.106em;">F</span> - <span class="mjx-char MJXc-TeX-math-I" style="padding-top: 0.472em; padding-bottom: 0.259em; padding-right: 0.106em;">(t,∞)</span> --- ## Generating the plotted points .panelset[ .panel[.panel-name[Overview] Regardless of how the axes are drawn, the plotted points are computed in a similar manner - X-axis: observed event times (failure, suspension, or other event) - Y-axis: nonparametric estimate of the failure probability (CDF) at the observed event times The nonparametric estimate of the CDF `\(\widehat{F}\)` plays a key role in probability plotting - One cannot observe the reliability or unreliability of a item - events are the only observable source of information - Reliability or unreliability values must be estimated from the data - The following panels discuss some considerations for computing the values of the plotted points ] .panel[.panel-name[X coordinates] .pull-left[ The x-coordinate of each plotted point is determined by how the observed event is categorized - Exact failure: "exact" failure time observed `\(t_f = t\)` - Left censored: failure is discovered at first inspection - exact failure time not known `\(t_f \in (0,t_{1})\)` - Interval censored: failure occurs between inspections and is discovered at inspection `\(i=2,\cdots,n\)` - exact failure time not known `\(t_f \in (t_{i},t_{i+1}), i > 0\)` - Right censored: failure not observed at final `\(n^{th}\)` inspection - exact failure time not known `\(t_f \in (t_{n},\infty\right)\)` Data sets that include only exact failures are called complete data sets ] .pull-right[ <img src="data:image/png;base64,#plotly_test_files/figure-html/unnamed-chunk-6-1.png" style="display: block; margin: auto;" /> ] ] .panel[.panel-name[Y coordinates] As previously stated, the y-coordinates of the plotted points are nonparametric estimates of the CDF corresponding to the observed event times Many different estimators have been developed for this purpose, some of which are - Plotting position estimators - Kaplan-Meier estimator (aka the product limit estimator) - Turnbull Estimator The choice of which estimator to use is driven by the type censoring (suspensions) that present in the data ] .panel[.panel-name[Plotting Positions] Probability plotting positions express the non-exceedance probability of the CDF for the `\(i^{th}\)` ascending data value The generic plotting position formula is expressed as $$ \widehat{F(t_{i})}=\frac{i-a}{n+1-2a} $$ - where + `\(i\)` is an index of the ordered observations (smallest `\(\rightarrow\)` largest) + `\(n\)` is the number of observations + `\(a\)` is the <span class="explain">plotting position parameter</span><span class="tooltip">The value of `\(a\)` is chosen to produce approximately unbiased estimates of `\(F(t_{i})\)` for an assumed distribution</span> Various formulae <span class="explain">have been developed</span> <span class="tooltip"> <br/> Author | Year | `\(a\)` | Formula -----------|------|-------|--------------- Hazen | 1914 | 0.50 | `\(\displaystyle\frac{i-0.5}{n}\)` Weibull | 1939 | 0 | `\(\displaystyle\frac{i}{n+1}\)` Blom | 1958 | 0.375 | `\(\displaystyle\frac{i-0.375}{n+0.25}\)` Gringorten | 1963 | 0.44 | `\(\displaystyle\frac{i-0.44}{n+0.12}\)` Chegodayev | 2000 | 0.30 | `\(\displaystyle\frac{i-0.30}{n+0.4}\)` Cunnane | 1977 | 0.40 | `\(\displaystyle\frac{i-0.40}{n+0.2}\)` Median | 1943 | 0.3175| `\(\displaystyle\frac{i-0.3175}{n+0.365}\)` <br/> The formulae used most often in practice are: <br/> $$ `\begin{aligned} \text{Hazen}:\; &\widehat{F(t_{i})}=\frac{i-0.5}{n}\\\\ \text{Chegodayev}:\; &\widehat{F(t_{i})}=\frac{i-0.3}{n+0.4}\\\\ \text{Weibull}:\; &\widehat{F(t_{i})}=\frac{i}{n+1}\\\\ \end{aligned}` $$ <br/> Note: the Chegodayev plotting position is also known as "Median-Ranks" </span> to correspond with specific distributions ] ] --- ## table Author | Year | `\(a\)` | Formula -----------|------|-------|--------------- Hazen | 1914 | 0.50 | `\(\displaystyle\frac{i-0.5}{n}\)` Weibull | 1939 | 0 | `\(\displaystyle\frac{i}{n+1}\)` Blom | 1958 | 0.375 | `\(\displaystyle\frac{i-0.375}{n+0.25}\)` Gringorten | 1963 | 0.44 | `\(\displaystyle\frac{i-0.44}{n+0.12}\)` Chegodayev | 2000 | 0.30 | `\(\displaystyle\frac{i-0.30}{n+0.4}\)` Cunnane | 1977 | 0.40 | `\(\displaystyle\frac{i-0.40}{n+0.2}\)` Median | 1943 | 0.3175| `\(\displaystyle\frac{i-0.3175}{n+0.365}\)` --- ## Distribution Functions - Quantile .panelset[.panel[.panel-name[Quantile Function] .pull-left[
] .pull-right[ The Quantile function (aka the percent point function) is formally defined as the realization of the R.V. <span class="mjx-char MJXc-TeX-math-I" style="display:inline;">T</span> that corresponds to the probability <span class="mjx-char MJXc-TeX-math-I" style="display:inline;">p</span> $$ t(p)=\inf\left\{t\in\mathbb{R}:p\leq\;F(t)\right\} $$ $$ t(p)=\inf \Big\\{t\in\mathbb{R}:p\leq\;F(t)\Big\\} $$ $$ t(p)=\inf \{t\in\mathbb{R} p\leq\;F(t)\} $$ ] ] ]